E. Qeli, W. Wiechert, and B. Freisleben (Germany)
visualization, sensitivity matrices, metabolic networks, multi-dimensional scaling.
Metabolic networks are an important research area in systems biology. To study their behavior, models are built and simulations are performed. Since such models depend on a possibly large number of parameters, sensitivity analysis is employed to investigate how sensitive a model is with respect to its parameters and how sensitive the estimated parameters are with respect to the experimental input data. The outputs of the sensitivity analysis are time varying matrices, which are difficult if not impossible to explore manually. In this paper, a novel approach to visualize time-varying sensitivity matrices generated during simulations of metabolic network models is presented. It is based on making use of multidimensional scaling, a dimension reduction technique, to extract similarity information. Implementation issues are discussed, and experimental results for a particular E.coli model are shown.
Important Links:
Go Back
